NZXT RESPONDS: ruclips.net/video/ZnDWxiFvUtE/видео.html Part 2 is up: ruclips.net/video/fjUscSRLwks/видео.html If you have one of these computers or cases, even if you have not experienced fire, please replace both screws immediately with something non-conductive. We would also personally advise replacing the PCIe riser in its entirety and getting rid of the included one in a way that it can't accidentally be used with a metal screw in the future. Please consider buying one of GN's Wireframe Mouse Mats to support us: IN STOCK & SHIPPING NOW! store.gamersnexus.net/products/gn-wireframe-mouse-mat - this is a high-quality desk-size mousing surface with custom blue rubber underside, blue stitched borders, and a high-quality print with PC component design. This is the best way to support our expensive, in-depth research pieces like this one.
The second Patrick to have been employed should have been required to change his name. Just as if two Matt's are employed, one must be renamed Gloss. If both go by Pat, the second should be called Stroke or Rub. Otherwise, Pat, Rick, or Paddy would be begrudgingly accepted.
Honestly, this is fairly wrong. The threads of the screw do not move relative to another stationary material. What was likely happening instead as he was manipulating the screw the board itself moved a little as clamping changed.
@@kevku2010 Axial movement of the 12V plane as the screw is tightened, _including via compression of the PCB,_ would be functionally equivalent to the movement shown in the render - on a much smaller scale, of course.
I suspect what’s happening is that you have to make just the right amount of short to get a bit of resistance but not so little that the power supply trips.
"This is the PCIeLawyer, and what I have for you today is a PCIe riser cable. I'm going to test it for shorts to grounds. Little beep out of one... little beep out of two..."
I'm electronics designer, and it's the basics of basics that every single screw hole on a PCB is either surrounded by ground, so that if anything ever happens, only thing that's going to create connection with the screw is ground... Or the area of the screw/mounting hole is completely vacant of any power planes so that there has to be a complete physical mishandling of the PCB to cause any issues. Someone wasn't thinking... someone *screwed*up...
Another electronics designer here and fully agree. There is probably a copper pour in that PCB which is meant to be connected up to GND but someone connected it to 12V rail instead. Sounds like there is no keepout around that screw at all which allowed this problem to manifest in the first place (and literally light up the screw up). I'd also suspect that moving the case around with a heavy graphics card in it might also cause a similar issue as the screw threads might be able to dig into the PCB more.
@@crenn6977 I guess they used a mirrored design of the cables other end, or something? since pin 1 is a 12v and the last pin is likely ground ? Maybe to avoid a 180 degree twist on the cable?
@@crenn6977 ya whenever he checked with ohm meter there was no connection from screw to ground and every time connection to 12v pci so they copper poured the 12v pci or even worse switched ground and 12v planes on the whole board which could cause noise issues
@@crenn6977 Honestly, as another fellow PCB designer, this is just... sad. It's such a *basic* thing that it's hard to believe such a major company would fuck it up, and even harder to believe that their best solution for fixing it was non-conductive screws. You're still exposing a power plane, even if you aren't actively shorting it yourself...
NZXT - "A very rare issue" Gamers Nexus - "A very serious rare issue" Drug Manufacturers - "In some very rare cases, patients experienced spontaneous combustion"
Thank god i haven't bought this case yet i still like it overall. Was going to do a compact Ryzen build w it. my 1st fire mistake as a techie was connecting a 5v pin of the same type to a 12v that caught the LED light on fire.
That poor little 5450 or whatever it was has dodged a bullet! To end up in the GN parts bin after such a long life and get through a GN video alive is amazing. I think we should give it an honorary spot on the wall! or maybe a cameo with Snowflake!
And that, my friends, is why when laying out a board you either have a ground plane or a keepout area around mounting regions. Also prevents broken traces if you have a gorilla mounting your hardware or your stuff gets subjected to excessive forces.
Yes, and what pisses me off the most is that this is automatic in PCB CAD software, you set a value for radius and that's pretty much it. The lack of effort on this is appalling.
Can we just appreciate the fact, that steve is not in any kind of way bashing nzxt but trying to find a serious solution for them just to protect people from a threatening fire.
The nylon screw workaround scares me. It should generally be safe, but it would be easy years down the line to forget that there was a reason the screw was nylon and replace it with a metal screw. The risk isn't worth it when you can end up facing the destructive and deadly force of an uncontrolled fire. I won't be purchasing anything from NZXT anymore. Not just because I'm wary of their engineering but also because they handled this poorly. The were not open about the exact issue and they went for the cheap fix instead of the right fix.
@@ziegfeld4131 you underestimated the human minds really. NZXT is a company large enough to be able to speculate that some will do exactly that. In a world where you need to put sticker "Caution, tempered glass" on the casing itself is self explanatory. Don't you think it requires a label in the cable "CAUTION ONLY REPLACE WITH PROVIDED NYLON SCREWS"? Especially in murica.
@@ziegfeld4131 it would be very easy for a second hand buyer to assume whoever sold the case just didn’t have a metal screw and grabbed whatever was laying around. It’s a simple and super plausible mistake. Hell I would probably think the same thing if I bought one second hand and discovered the screw, or didn’t notice it when disassembling it, then when rebuilding it either put the nylon screw in the top screw hole or just swapped it for a metal one. It’s not a stupid mistake to make. Shame on NZXT
There's no other solution than recall that riser card. Ditch it and don't look back, put a proper one in place, that riser is ridiculous and that solution is appalling. The nylon screw is just the cheapest solution, it's absolutely ridiculous. Can't understand how anyone with a PCB CAD software could do something so stupid on that riser without realizing it. It's just setting clearances on screw mount holes, it's automatic even. I doubt anyone now would still have this problem but no doubt, if you're getting this case or pre-built second hand, ditch the riser card and get a good one in it's place.
@@Wentr-z It really is very easy. It would never have even crossed my mind that it might have something to do with fire-risks if I bought a case second-hand and the screws were nylon. Not before this entire debacle came to my attention. As naive as it sounds, a lot of people would just assume that a known brand wouldn't make these basic mistakes out of cheapness. Why? Because most cases don't tend to be fire-risks when using normal damn screws.
What really surprises me here is how this could even have occurred in the first place. All PCB programs throw big red design rule check errors when you put a trace near a screw. And any PCB houses fabrication rule checks would flag this immediately too for violating clearance rules. So I'm really not sure how something like this can even get to the point of shipping in the first place.
@@soccerguy2433 Yeah, and seeing how difficult this is to replicate it wouldn't come up in testing. NZXT is ultimately responsible, but they have probably just ordered the PCB and expected them to be according to spec. It's also possible some issue has slipped through in the design, software checks aren't perfect either.
They most likely have a tolerance in screw hole placement, but the trace is too close anyway. The PCBs my company makes for our analytical instruments all have a mounting border to prevent this entirely.
just dont buy NZXT, they are fucking trash. I ordered a 5pack of fans couple of years ago and even that they cant get right. 5 rattling fans out of the box... If they cant even make fans work decently, i'd be worried about any electrical component. Btw, dont get their AIO watercoolers unless you like a sprinkler in your case.
Lathan has been on quite a few times, but mostly streams. Patrick Stone has also had many appearances in years past, but it's been a while since he wasn't formally a part of the team until this month!
@@GamersNexus If it's feasible, I'd love seeing them on for future normal videos; gotta love the drawings and the variety absolutely helps make the content more engaging. I was surprised how much I liked having all three on at once!
Patrick be like: "This is The Shockpicking Lawyer, and what I have you today is a faulty raiser cable. Nothing on 1. Nothing on 2 or 3. Beep out of 4.... "
That short circuit that has enough resistance to keep the circuit energized is simply called a high resistance short and as you just demonstrated are dangerous as hell.
Very impressive troubleshooting guys! From a life long career of working in the electronics manufacturing as well as being a repair technician I can estimate a LOT of time and effort going into this video that didn't make it into the finale product. From my experience things like this can easily take months if not years to find and correct. I just hope NZXT appreciates what you have done for them and takes advantage of this effort to resolve the design flaw instead of taking a defensive position and try to pass the blame back to your efforts as I have seen big business do in the past. Love your channel, your ethics, your approach and your crew. Thank you.
I heard a rumour that employment application forms for GN already have the "First Name" field populated with Patrick to save any future applicants time.
There's a third factor beyond probability of occurrence and severity: probability of detection. In this case, other than a fire (the failed state), there's no reasonable way a user would know that they have a high risk of failure, which makes this problem much worse than when NZXT tried to downplay it with the "low occurrence" aspect. (See FMEA for more info).
i really enjoy how he speaks. maybe its a trade thing. a guy who knows what he is talking about and lays it out clearly for everyone to understand. for some reason the comments wont load on the follow up video. its sad this type of thing can get a channel or comment section banned. i cant think of a more serious problem/defect for a company like nzxt.
I'm puzzled to why they would design a pcb that way, even motherboards with more complex wiring and circuits gives at least 3.5mm radius clearance on screw holes, just grounding.
As an electrical engineer and PCB designer some one messed up with the design rules for clearance on grounded fixings ...even Free PCB software like kicad has design rule tools that would have prevented this unless the PCB designer deliberately ignored the warnings that pop up when running the design rule check....the PCB risers are junk and should be replaced by NZXT with a new design immediately... I'd be ashamed to leave a product like that in use in the public domain...... excellent work GN Team on the video.
so if Patrick and Patrick couldn't get results, we'd turn the whole thing over to Patrick for further testing. Barring that we can let Patrick have a shot at it.
This content certainly provides a lot of diagnostics that could be useful to them, so we hope that the company is able to gather some useful information and prevent issues going forward.
lol. He does his job and he does it well, the facts and reasoning of e v e r y t h i n g pc related. You make a dodgy product/make a mistake, you get a low rating. But if you do your job well and the product works as intended, ur good in hood. But yeah ofcourse they will hate him when they try and cover mistakes, many large companies and organizations have a love hate relationship with the media. While on one hand they can significantly help spreading your brand and basically being overpowered advertising, they can also cripple your reputation just by one mistake you do.
They should've combined this case with the use of the PowerColor Red Dragon Radeon RX 5600 XT (known as the card that overheats badly in the GN review).
@@jackmiller8851 it’s just for this video. It so happens that they have a few people on staff named Patrick who also happen to be in the video, so they made a meme about it.
You jest, buuut I'll be including some insulation in my case as of today. If a company like NZXT doesn't give a shit, what can I expect from "lesser" brands?
@@dirtycopgangstaoriginal6971 You have a pci-e riser in there? If not you likely don't need to. The issue isn't the chassis, the chassis _should be_ grounded/in contact to the psu. The issue here is the pci-e riser having both an issue in the wiring of the power pins and the ground plane of the pcb having too little safety margin(if any) around the screw mounting holes.
I just removed the screws and replaced them with two zip ties. Different issue all together the components run super hot anyway. In the UK the replacement screw kits are delayed because of BREXIT and the national lockdown as well. Yes, they are sending out nylon screws, however zip ties have help up just fine as well.
@@GamersNexus Yes, I replaced both. Tightened them well and cut the tails with a tiny margin left on each side. It has worked fine up till now. I changed it around Christmas and have not had any issues using my PC like this. Thanks for the content and cheers!
Interesting, the lock down affecting traffic within UK make sense, but Brexit come into play? I thought cross channel shipping continues as normal, well, except the additional checking due to the suppose new Covid mutation in UK.
@@Verpal Brexit has caused massive delays because whole new shipping protocols and customs checks have to be implemented. Ireland, being Western Europe getting most things through the Uk is also experiencing massive delays in all shipments including supermarkets
@@dalyo96 Oh! I thought you are living in UK, if you are living in Ireland indeed Brexit will cause trouble, especially when UK seems to prioritize sorting border and shipping with continental Europe.
Shocked that this actually made it to market. I'm not a fan of the case but as an ITX fan I was super pleased to see mainstream manufacturers bringing ITX to the masses
NZXT -"This is our special edition airflow detector case. It automatically generates white smoke so you can see where the air is going" *Disclaimer - We guarantee this to happen once. Make sure you have insurance.
I can't imagine why there would be a 12v trace just happening to run past that mounting hole - there's nothing out there. I bet that's just a flood filled 12v power plane spanning the entire layer of the PCB. Which is fine, but why isn't there any pullback around the mounting holes? Did the layout guy just forget to put the mounting holes in the layout? Did they add mounting holes after layout was done, with exposed 12v power plane all around the inside of the mounting hole? Would be real interesting to see a copy of the board file for that thing.
This^^ is all the explanation needed, succinct and didn't need 3 hours of rambling nonsense. A dead short to ground because the 12v copper was right up on the hole. Suggesting they should have plated the hole was hilarious. That would result in a more "perfect" short to the screw, making a fault even more likely.
In part two thetes a visual where the 12v circuit is touching the screws which the screws which touches the case plate which serves as part of ground. Just really bad circuit diagram design, underestimating pcb wear from screwing at unprecise angles (which is 99.9%) of how we typically screw something with hand. The metal screw threads off parts of thr pcb and exposing the 12v circuit unto the screws. Just sad that even motherboards with more complex wiring and circuits managed to squeeze a proper screw hole with plating and gaps and this riser board couldn't.
great footwork and deductive reasoning on the process that wound up in a case of fault... showing the when and how a failure can happen and what that failure looks like is definitely an educational resource to the masses.
This needs more upvotes. Looks like NZXT patched the problem with the cheapest option available instead of fixing it by recalling riser card and replacing it with new design. In the end problem is pushed to customer, and other hardware manufacturers that will have to handle RMA’s from shorted Graphics cards etc.
You, or someone else who might use it.. edit: also, the ones who so far haven't had issues may experience them just by moving the PC around a bit, where the moving GPU might slightly loosen the screw and cause it to short
You know, i'd drill out the hole larger and then fill it with epoxy and drill back out to screw size - the PCB material is epoxy-glassfibre to begin with. So the board would essentially be guaranteed to spec then. There can't be anything actually useful in that part of the board, probably just an overhang of a layer pour. But then that's me, i'm a crazy person. They really should remanufacture the boards and mail out replacements, it's not all that expensive. But it does take time, a few months, which is something they don't have at the moment. You can get these nylon screws at moment's notice from numerous suppliers around the world, so i can't exactly blame them for taking immediate action and not sitting on their arse. Maybe they could do a second pass and now also send everyone a new riser a month or two later. AND MARK IT SOMEHOW so people don't confuse them and can determine whether they have a good riser in their PC or not.
I remember years ago about the time that the DX4-100 was released when a machine came into the shop from a distressed end user that had caught fire in an unusual way, 2 wires had burned the insulation off in the centre of the floppy drive ribbon from the Vesa IO card to the back of the floppy drive, the cable seemed to be the only casualty - on inspection it was discovered that the end user missed a pin on the smaller molex power connector when refitting after carrying out an upgrade, even though it was the data cable that burned - strangely enough the IO card and the floppy drive survived -
Steve and company, great video, love the work you do bringing the techie side of computers to the masses. I do have one issue with one assumption regarding the screw placement and the motion of the thread on the screw. The thread itself would not move up and down as shown in the video. Simply look at the inside of a nut and notice it would be in a fixed position, same goes for the thread on the screw as it moves in and out of the screw-hole. A possible reason why some systems are affected and not others maybe as a unit is assembled the screw could be inserted in such a way that the pitch of the threads catches the trace on the PCB. One could assume any number of possible ways a screw can be inserted may result in any number of possible ways the thread catches the PCB trace. I do think it kind of sloppy that the PCB was designed in such a way that the result can be so dangerous. Anyhow, keep up the good work!!
"had this for 5 years, surprised that it still works" he says referring to the "fire extinguisher" as the fire is already started and he is trying to put it out lol
They have actual serious fire extinguishers too. They took the little spray bottle extinguisher because it doesn't require a lot of cleanup and is cheap to replenish. And because it was due for replacement anyway.
Yeah, I have real fire extinguishers as part of standard office requirements. I just didn't want to use one that important for a fire I had full control of. I mean, we could have just thrown the case in the parking lot if we needed to, or picked up the real fire extinguisher.
Sorry Steve.... when i first saw the Vid/Thumbnail I assumed it to be clickbait - since this is YT after all. Coming across the story now I realise I should have never doubted your integrity. Thank you for keeping it real.
In fact, your PCB software will yell at you, your fab will yell at you and the test engineers will yell at you. I don't know how something like this can make it into production.
@@squelchedotter Unless, big brain moment in here... You don't add any mounting holes in the design soft since your customer was like, no need, no need - we sort it out ourselves...
@@squelchedotter as a last minute change that bypasses most of the people and machinery in the chain for example. You can easily edit the Exellon drill file by hand.
@@konzetsu6068 Best practices is to leave a keep out zone for all layers. For exactly this type of problem. Even KiCad will do this automatically unless you override the settings.
So the thing is with screws, the thread doesn't move up or down at a particular connection point. It slides in the slot, but that slot remains constant. The CGI video you show is not accurate of how a screw moves into it's threaded hole. There is some movement, technically, on an extremely small scale that is probably indiscernible to the unaided human eye. i will agree that the minute movement of simply turning the screw triggers the short, as any pressure moves the screw in an incredibly small amount up or down in the hole. The original screw is painted so not only does the screw need to move (and not just rotate) in the hole but that has to coincide with stripping of paint off that screw. It also appears, to me, that a layer inside the PCB burns up, through the entirety of the PCB. The fire itself seems to be a byproduct of the short where the plastic catches fire, but I'm fairly certain that that particular layer of material inside the PCB burns (albeit with no flames, but things can burn without flame).
I... think this might be the episode to show to my dad, the retired electrical engineer who worked his way up to become a global manager for risk management of medical devices at a certain large safety science certification company. He'll appreciate it! :D
NZXT RESPONDS: ruclips.net/video/ZnDWxiFvUtE/видео.html
Part 2 is up: ruclips.net/video/fjUscSRLwks/видео.html
If you have one of these computers or cases, even if you have not experienced fire, please replace both screws immediately with something non-conductive. We would also personally advise replacing the PCIe riser in its entirety and getting rid of the included one in a way that it can't accidentally be used with a metal screw in the future.
Please consider buying one of GN's Wireframe Mouse Mats to support us: IN STOCK & SHIPPING NOW! store.gamersnexus.net/products/gn-wireframe-mouse-mat - this is a high-quality desk-size mousing surface with custom blue rubber underside, blue stitched borders, and a high-quality print with PC component design. This is the best way to support our expensive, in-depth research pieces like this one.
this is some shitty design consideration on NZXT's part ngl
NZXT is an shitty company
@@PlayerSlotAvailable FOLLOW STEVE'S ADVICE!
did you use a different screw to mount that riser? wasn it black and not flared?
could have been someone during assembly that used the wrong screw and not a spacer problem.
Serous issue if the Patrick-in-Chief AND the Chief Patrick officer got involved!😎
Patrick Patrick Patrick Patrick
Patrick Patrick Patrick Patrick Patrick Patrick Patrick Patrick
The second Patrick to have been employed should have been required to change his name. Just as if two Matt's are employed, one must be renamed Gloss. If both go by Pat, the second should be called Stroke or Rub. Otherwise, Pat, Rick, or Paddy would be begrudgingly accepted.
@@javadkhusro Its the gamers nexus hiring policy:)
@@poderr ok that really made me lol
If you have a kickass last name like "Stone" that should be what people call you. :)
Awesome video but the best part was dragging the cable through the thermal paste at 4:45
Ahaha sloppy work Steve
I lost my shit when he did that! He does it on purpose lmao
Of course der8auer would have his thermal paste senses go off from that!
I saw it too lol, but I figured since they are trying to start a campfire with a computer he didn't care much lol
Gave the cable a little sample.
The 3D render of the contact inside the PCB was really effective at showing the problem. More of this sort thing, please.
Honestly, this is fairly wrong. The threads of the screw do not move relative to another stationary material.
What was likely happening instead as he was manipulating the screw the board itself moved a little as clamping changed.
@@kevku2010 You're right, unless they were showing what happens when it's stripped out. The world may never know.
@@kevku2010 Axial movement of the 12V plane as the screw is tightened, _including via compression of the PCB,_ would be functionally equivalent to the movement shown in the render - on a much smaller scale, of course.
I suspect what’s happening is that you have to make just the right amount of short to get a bit of resistance but not so little that the power supply trips.
simple 3d models can explain so many things so well
PCIe riser : catches fire
GN : "keep it rolling"
The motherboard just behind the tray : "this is fine"
"This is the PCIeLawyer, and what I have for you today is a PCIe riser cable. I'm going to test it for shorts to grounds. Little beep out of one... little beep out of two..."
We have a person of culture right here.
Thought the same thing!
I understood that reference!
Yes, exactly what I was thinking 😂
Must be in the same category as dentists..
13:38 All I can hear here is Lock Picking Lawyer "Nothing on one, two's binding, click out of three..."
i was looking for this comment
Wait, Steve's a LPL fan?! We need a crossover!
Oh good god, not just me.
Did the same exact thing
Love how literally EVERYONE knows who the LPL is lol
"What's the reading on the thermocouple?" "FIRE"
This would be a good meme for Fire Marshal Bill.
@@antc1442 I see, you're a man of culture as well. Noice.
It's pretty lit.
I believe the correct reading is "Holy shit! I'm on fire! Ow! Ow!! Ow!"
Will Smith: Oh, that's hot.
I'm electronics designer, and it's the basics of basics that every single screw hole on a PCB is either surrounded by ground, so that if anything ever happens, only thing that's going to create connection with the screw is ground...
Or the area of the screw/mounting hole is completely vacant of any power planes so that there has to be a complete physical mishandling of the PCB to cause any issues.
Someone wasn't thinking... someone *screwed*up...
Another electronics designer here and fully agree. There is probably a copper pour in that PCB which is meant to be connected up to GND but someone connected it to 12V rail instead. Sounds like there is no keepout around that screw at all which allowed this problem to manifest in the first place (and literally light up the screw up). I'd also suspect that moving the case around with a heavy graphics card in it might also cause a similar issue as the screw threads might be able to dig into the PCB more.
ba dum tsss
@@crenn6977 I guess they used a mirrored design of the cables other end, or something? since pin 1 is a 12v and the last pin is likely ground ? Maybe to avoid a 180 degree twist on the cable?
@@crenn6977 ya whenever he checked with ohm meter there was no connection from screw to ground and every time connection to 12v pci so they copper poured the 12v pci or even worse switched ground and 12v planes on the whole board which could cause noise issues
@@crenn6977 Honestly, as another fellow PCB designer, this is just... sad. It's such a *basic* thing that it's hard to believe such a major company would fuck it up, and even harder to believe that their best solution for fixing it was non-conductive screws. You're still exposing a power plane, even if you aren't actively shorting it yourself...
"fewer than 10"
That caught fire yes, i bet there is a lot of people that think their psu is broken because it keeps tripping
Winner of understatement of 2021 goes to:
"Housefires have actual real world consequences."
LOL
NZXT - "A very rare issue"
Gamers Nexus - "A very serious rare issue"
Drug Manufacturers - "In some very rare cases, patients experienced spontaneous combustion"
Cool kids: "Shit was lit yo"
Its a feature some would say
Only 1 of 1000 houses are burning down. 😁
Comfirmed: Steve is useless at remembering names so hires as many Patrick's as possible
just rename them officially when they sign the contract.
Patrick's Nexus
George foreman did...
Steve ain't Linux.
I don't blame him after he tried reading modern Intel mobile CPU names.
"You get to live another day"
Graphics Card: *heavy breathing*
Still better than getting enslaved for 24/7 Ethereum mining.
@@drinkingguy3168 I would have nothing but respect for anyone attempting to mine anything on a hd 5450
Thank god i haven't bought this case yet i still like it overall. Was going to do a compact Ryzen build w it. my 1st fire mistake as a techie was connecting a 5v pin of the same type to a 12v that caught the LED light on fire.
That poor little 5450 or whatever it was has dodged a bullet! To end up in the GN parts bin after such a long life and get through a GN video alive is amazing. I think we should give it an honorary spot on the wall! or maybe a cameo with Snowflake!
And that, my friends, is why when laying out a board you either have a ground plane or a keepout area around mounting regions. Also prevents broken traces if you have a gorilla mounting your hardware or your stuff gets subjected to excessive forces.
Yes, and what pisses me off the most is that this is automatic in PCB CAD software, you set a value for radius and that's pretty much it. The lack of effort on this is appalling.
Can we just appreciate the fact, that steve is not in any kind of way bashing nzxt but trying to find a serious solution for them just to protect people from a threatening fire.
6:30 "it didn't catch fire, but we're committed". Mythbusters vibe intensifies
"Conduct the fire to the heatsink" - NZXT
"Just pull hard on the fire, it will go right out"
Hello, Is this Gamers Nexus?
No, this is Patrick.
I was wondering how many comments I would have to go down till I found a Sponge Bob reference! Good Job man!
Patrick Star, It Burns!!!!!!
Is this Patrick?
No, this is Patrick.
Me reading this comment: ruclips.net/video/4r18Ab8cb5s/видео.html
@@CobraDBlade lol yes!
As a fellow Patrick, I am proud that the Patrick Industrial Complex is working nominally. Continue on.
The nylon screw workaround scares me. It should generally be safe, but it would be easy years down the line to forget that there was a reason the screw was nylon and replace it with a metal screw. The risk isn't worth it when you can end up facing the destructive and deadly force of an uncontrolled fire.
I won't be purchasing anything from NZXT anymore. Not just because I'm wary of their engineering but also because they handled this poorly. The were not open about the exact issue and they went for the cheap fix instead of the right fix.
If you replace a nylon screw with a metal 1 on anything youre a moron you should always replace stuff with stuff thats made the same
@@ziegfeld4131 you underestimated the human minds really.
NZXT is a company large enough to be able to speculate that some will do exactly that. In a world where you need to put sticker "Caution, tempered glass" on the casing itself is self explanatory.
Don't you think it requires a label in the cable "CAUTION ONLY REPLACE WITH PROVIDED NYLON SCREWS"? Especially in murica.
@@ziegfeld4131 it would be very easy for a second hand buyer to assume whoever sold the case just didn’t have a metal screw and grabbed whatever was laying around. It’s a simple and super plausible mistake. Hell I would probably think the same thing if I bought one second hand and discovered the screw, or didn’t notice it when disassembling it, then when rebuilding it either put the nylon screw in the top screw hole or just swapped it for a metal one. It’s not a stupid mistake to make. Shame on NZXT
There's no other solution than recall that riser card. Ditch it and don't look back, put a proper one in place, that riser is ridiculous and that solution is appalling. The nylon screw is just the cheapest solution, it's absolutely ridiculous. Can't understand how anyone with a PCB CAD software could do something so stupid on that riser without realizing it. It's just setting clearances on screw mount holes, it's automatic even.
I doubt anyone now would still have this problem but no doubt, if you're getting this case or pre-built second hand, ditch the riser card and get a good one in it's place.
@@Wentr-z It really is very easy. It would never have even crossed my mind that it might have something to do with fire-risks if I bought a case second-hand and the screws were nylon. Not before this entire debacle came to my attention. As naive as it sounds, a lot of people would just assume that a known brand wouldn't make these basic mistakes out of cheapness. Why? Because most cases don't tend to be fire-risks when using normal damn screws.
Did a great job of explaining and illustrating the issue.
Thanks! Patrick(s) are good at that stuff.
@@GamersNexus Patricks in general or just those 2?
@@Ikxi It's a small sample size, but so far 100% of the Patricks tested are good at it. :)
This problem its in ALL small pcb who use 12V and bend at heavy loads(gpu) .
@@HMarcBower These are obviously binned Patricks.
GN: The place that answers all the burning questions.
Ba Dum Tiss
😄
What really surprises me here is how this could even have occurred in the first place.
All PCB programs throw big red design rule check errors when you put a trace near a screw. And any PCB houses fabrication rule checks would flag this immediately too for violating clearance rules. So I'm really not sure how something like this can even get to the point of shipping in the first place.
That's a really good point
Wouldn't surprise me if it was a communication issue between departments/outsourced staff
so it checks on the program but perhaps isn't manufactured according to design?
@@soccerguy2433 Yeah, and seeing how difficult this is to replicate it wouldn't come up in testing. NZXT is ultimately responsible, but they have probably just ordered the PCB and expected them to be according to spec. It's also possible some issue has slipped through in the design, software checks aren't perfect either.
They most likely have a tolerance in screw hole placement, but the trace is too close anyway. The PCBs my company makes for our analytical instruments all have a mounting border to prevent this entirely.
13:39 lockpickinglawyer would be proud of Patrick
Yes I was thinking the same thing.
@@sweetnessinc ahah ikr
This was way more technical than I expected I love it
I received my nylon screws from NZXT a few weeks ago, but this video convinced me to finally install them.
replace the riser as well to completely mitigate it and eliminate the potential (lol, electrical joke right there...)...
I agree replace the cable. And cut that one at the board. Unless you need something to start your bon fire.
I'd be changing out that whole riser for one that is designed and grounded properly. That's a better solution!
just dont buy NZXT, they are fucking trash.
I ordered a 5pack of fans couple of years ago and even that they cant get right. 5 rattling fans out of the box...
If they cant even make fans work decently, i'd be worried about any electrical component.
Btw, dont get their AIO watercoolers unless you like a sprinkler in your case.
congrats to the patricks on speaking on camera for the first time. I don't mind seeing more of the team.
Lathan has been on quite a few times, but mostly streams. Patrick Stone has also had many appearances in years past, but it's been a while since he wasn't formally a part of the team until this month!
@@GamersNexus If it's feasible, I'd love seeing them on for future normal videos; gotta love the drawings and the variety absolutely helps make the content more engaging. I was surprised how much I liked having all three on at once!
I’ve seen Patrick on camera in several videos but this is the first video I’ve seen Patrick in.
Patrick be like:
"This is The Shockpicking Lawyer, and what I have you today is a faulty raiser cable. Nothing on 1. Nothing on 2 or 3. Beep out of 4.... "
That short circuit that has enough resistance to keep the circuit energized is simply called a high resistance short and as you just demonstrated are dangerous as hell.
24:20
pc on fire: "Do we have enough footage you think?"
dedication xD
Nice screensaver
@@tj71520 it's a hot commodity :P.
@@digitalsparky yes. Just the thing to build a hot pc
13:37 "Nothing on one"
"...two is binding"
1337! Sorry..
TLPL
With a small pin shape object on hand.
"Beep on three"
I don’t get it
When he said "nothing on one" I instinctively followed up with "two is binding." Too much LPL for me
Very impressive troubleshooting guys! From a life long career of working in the electronics manufacturing as well as being a repair technician I can estimate a LOT of time and effort going into this video that didn't make it into the finale product. From my experience things like this can easily take months if not years to find and correct. I just hope NZXT appreciates what you have done for them and takes advantage of this effort to resolve the design flaw instead of taking a defensive position and try to pass the blame back to your efforts as I have seen big business do in the past.
Love your channel, your ethics, your approach and your crew.
Thank you.
“Hello, is this Patrick?”
“No, this is Patrick”
and This is Patrick from STH ... :P
Insert Patrick Stewart
Which one is the "chief Patrick officer" ?
NZXT "we have a nylon screw"
GN "oh yeah, well we've got two Patricks"
As a Patrick, I approve of the ratio of Patricks working in your office!
So they probably have one going by Pat and the other by Patrick. I pity the Patrick that gets nicknamed Patty.
I'm also a Patrick, I'm glad to know there's at least one workplace which has a quota on Patricks
13:38
Patrick: 𝑛𝑜𝑡ℎ𝑖𝑛𝑔 𝑜𝑛 𝑜𝑛𝑒, 𝑡𝑤𝑜, 𝑡ℎ𝑟𝑒𝑒
LockPickingLawyer: 𝑾𝑨𝑰𝑻. 𝑻𝑯𝑨𝑻'𝑺 𝑰𝑳𝑳𝑬𝑮𝑨𝑳
Haha as soon as he said nothing on 4 I suddenly thought I was watching lpl haha
5 is binding
Nice click on three. Wait, not that video.
Thanks!
27:36 Gamers Nexus would not be Gamers Nexus if there would not be at least one chart in the video
And atleast one Patrick
Chart just to provide figures to prove that the fire is, in fact, a fire.
Hmm yes, fire is made of hot.
If Steve isn't Patrick, is Patrick then not Steve?
Sir, this isn't a philosophy channel!
@@GamersNexus Is this the Krusty Krab?
Think it's time for a Ask GN episode, entirely dedicated to this question!
A Steve can be a Patrick, but a Patrick can't be a Steve. Kinda like squares and rectangles.
NO! Patrick is Patrick.
I heard a rumour that employment application forms for GN already have the "First Name" field populated with Patrick to save any future applicants time.
This has to be one of the best investigative pieces you have done! Thanks so much to you and your team for shedding more light on this situation.
The content is great with proper testing which can't be appreciated enough. The cherry on the top is the humor. Raise my glass to you guys.
You know it's good when it starts with a fire and a "Not A Patrick" name tag for Steve 🤣
15:02 Someone had fun there 🤣
There's a third factor beyond probability of occurrence and severity: probability of detection. In this case, other than a fire (the failed state), there's no reasonable way a user would know that they have a high risk of failure, which makes this problem much worse than when NZXT tried to downplay it with the "low occurrence" aspect. (See FMEA for more info).
*When the video starts off with "ohhh shit", you know it's going to be GOOD...* 🔥
stfu
@@pandasdreamlygon why?
@@moment3380 coz someone needs attention, duh!
As a former electrical engineer this is like the best porn. :p
Great design NZXT ;)
i really enjoy how he speaks. maybe its a trade thing. a guy who knows what he is talking about and lays it out clearly for everyone to understand. for some reason the comments wont load on the follow up video. its sad this type of thing can get a channel or comment section banned. i cant think of a more serious problem/defect for a company like nzxt.
I'm puzzled to why they would design a pcb that way, even motherboards with more complex wiring and circuits gives at least 3.5mm radius clearance on screw holes, just grounding.
As an electrical engineer and PCB designer some one messed up with the design rules for clearance on grounded fixings ...even Free PCB software like kicad has design rule tools that would have prevented this unless the PCB designer deliberately ignored the warnings that pop up when running the design rule check....the PCB risers are junk and should be replaced by NZXT with a new design immediately... I'd be ashamed to leave a product like that in use in the public domain...... excellent work GN Team on the video.
so if Patrick and Patrick couldn't get results, we'd turn the whole thing over to Patrick for further testing. Barring that we can let Patrick have a shot at it.
Avoid Patrick. Patrick says that guy is incompetent, you should get Patrick or Patrick on the job. Patrick is pretty good though, he should sort it.
We could call them all Bruce, that was proven to solve issues .... and now for something completely different ..... 🐍
I wonder how fond NZXT is of Steve
This content certainly provides a lot of diagnostics that could be useful to them, so we hope that the company is able to gather some useful information and prevent issues going forward.
Puck GamersNexus 😋
lol. He does his job and he does it well, the facts and reasoning of e v e r y t h i n g pc related. You make a dodgy product/make a mistake, you get a low rating. But if you do your job well and the product works as intended, ur good in hood. But yeah ofcourse they will hate him when they try and cover mistakes, many large companies and organizations have a love hate relationship with the media. While on one hand they can significantly help spreading your brand and basically being overpowered advertising, they can also cripple your reputation just by one mistake you do.
If NZXT wanted them to focus on current tracing then YES...
@@GamersNexus considering the are a big company they will probably learn nothing
NZXT H1 Case Fire Thermal Testing | Answering: "Is Fire Hot?" lmao
Only if it's NZXT branded Fire™ apparently
As someone who used to play a healer in WoW I can tell you a surprising number of gamers are unaware that fire is actually hot.
Why is it a different screw?
Still, not hot enough to melt steel beams. ;)
2:30 was the most thorough explanation I've ever heard regarding why computers should not be on fire.
"Power supplies can certainly cause fires"
If only Steve from the past knew what the future has prepared for him...
fire exclamation mark fire exclamation mark
"The IT Crowd" :)
I'm so glad I wasn't the only one who thought of that episode
They should've combined this case with the use of the PowerColor Red Dragon Radeon RX 5600 XT (known as the card that overheats badly in the GN review).
Remember the new number!
0118 999 881 999 119 725
3
Looking forward to hearing from you !
This was exactly what this channel needed: Electricity, fire and Patricks
I feel like Patrick-in-Chief should be a thing in every company.
Sorry ive been away a while, what's with the patrick thing?
@@jackmiller8851 it’s just for this video. It so happens that they have a few people on staff named Patrick who also happen to be in the video, so they made a meme about it.
"I think this riser cable's bad, someone throw me a Patrick."
"Power supplies certainly can cause fires"... It's like he's clairvoyant!
"Once you introduce fire to a computer, things will start dying." - Steve 2021
So I need rubber pads to insulate my power supply
-The Verge
LMAO
HAHA! They knew! They were way ahead of the curve. rofl
Yeah, it's not as if _the case is supposed to ground to the PSU..._
You jest, buuut I'll be including some insulation in my case as of today. If a company like NZXT doesn't give a shit, what can I expect from "lesser" brands?
@@dirtycopgangstaoriginal6971 You have a pci-e riser in there? If not you likely don't need to. The issue isn't the chassis, the chassis _should be_ grounded/in contact to the psu.
The issue here is the pci-e riser having both an issue in the wiring of the power pins and the ground plane of the pcb having too little safety margin(if any) around the screw mounting holes.
less than a month in and we already have our 2021 disappointment case
Thats achievement not to brag about. SERIOUS problem.
I really wanted an H1. Now after this video, would never even consider it. Very well done and informative work on this video.
“And now we’re gonna try and light it on fire” one of the best lines in a computer case video ever 😂
I just removed the screws and replaced them with two zip ties. Different issue all together the components run super hot anyway. In the UK the replacement screw kits are delayed because of BREXIT and the national lockdown as well. Yes, they are sending out nylon screws, however zip ties have help up just fine as well.
As long as you replaced both, from what we understand now, it seems like that'd be an adequate work-around to the issue.
@@GamersNexus Yes, I replaced both. Tightened them well and cut the tails with a tiny margin left on each side. It has worked fine up till now. I changed it around Christmas and have not had any issues using my PC like this. Thanks for the content and cheers!
Interesting, the lock down affecting traffic within UK make sense, but Brexit come into play? I thought cross channel shipping continues as normal, well, except the additional checking due to the suppose new Covid mutation in UK.
@@Verpal Brexit has caused massive delays because whole new shipping protocols and customs checks have to be implemented. Ireland, being Western Europe getting most things through the Uk is also experiencing massive delays in all shipments including supermarkets
@@dalyo96 Oh! I thought you are living in UK, if you are living in Ireland indeed Brexit will cause trouble, especially when UK seems to prioritize sorting border and shipping with continental Europe.
Shocked that this actually made it to market. I'm not a fan of the case but as an ITX fan I was super pleased to see mainstream manufacturers bringing ITX to the masses
“This screw is happening” - Quote of how 2020 was
Timestamp pls
As an engineer, I appreciate the thorough explanations and attempts! Great vídeo, very educating.
"and now we're gonna try to light it on fire"... You have earned my like and subscribe
Man, this case is FIRE
_Underrated_ _Comment_
Bro, this case is lit
@@thevaultdwellergaming i was about to comment that oof
Like literally, _dawg!_
NZXT -"This is our special edition airflow detector case. It automatically generates white smoke so you can see where the air is going" *Disclaimer - We guarantee this to happen once. Make sure you have insurance.
Nzxt coming in hot with yet another great airflow case
Steve:Oooh, fire, better get that. Watches it burn.
I can't imagine why there would be a 12v trace just happening to run past that mounting hole - there's nothing out there. I bet that's just a flood filled 12v power plane spanning the entire layer of the PCB. Which is fine, but why isn't there any pullback around the mounting holes? Did the layout guy just forget to put the mounting holes in the layout? Did they add mounting holes after layout was done, with exposed 12v power plane all around the inside of the mounting hole?
Would be real interesting to see a copy of the board file for that thing.
Don't need a schematic if you can just dremel your way through the layers
That's what I thought. Even the most cheapest of riser cables don't typically have this issue.
I guess the PCB layout was never properly peer-reviewed. Whoever designed it should be fired. It's shameful. I designed better PCB 15 years ago.
This^^ is all the explanation needed, succinct and didn't need 3 hours of rambling nonsense. A dead short to ground because the 12v copper was right up on the hole. Suggesting they should have plated the hole was hilarious. That would result in a more "perfect" short to the screw, making a fault even more likely.
In part two thetes a visual where the 12v circuit is touching the screws which the screws which touches the case plate which serves as part of ground.
Just really bad circuit diagram design, underestimating pcb wear from screwing at unprecise angles (which is 99.9%) of how we typically screw something with hand. The metal screw threads off parts of thr pcb and exposing the 12v circuit unto the screws.
Just sad that even motherboards with more complex wiring and circuits managed to squeeze a proper screw hole with plating and gaps and this riser board couldn't.
I wonder if Verge guy is watching and thinking "see, I told you the PSU needs insulators". 😜
LOOL
Cave men : make fire with stones and sticks
Modern men : make fire with PC components
I like patrick. He has AvE/This old Tony vibes. Especially with that camera angle
great footwork and deductive reasoning on the process that wound up in a case of fault... showing the when and how a failure can happen and what that failure looks like is definitely an educational resource to the masses.
Amazing job figuring out the issues with the H1, and even more amazing is how you explained it in such simple terms. Keep up the great work GN crew!
"Before that"
Gotta screw with confidence Steve
I love you gang
eventually, you'll forget that this particular riser can't use metal screws, because why the hell wouldn't it? Cut it in half and throw it away.
This needs more upvotes. Looks like NZXT patched the problem with the cheapest option available instead of fixing it by recalling riser card and replacing it with new design. In the end problem is pushed to customer, and other hardware manufacturers that will have to handle RMA’s from shorted Graphics cards etc.
You, or someone else who might use it..
edit: also, the ones who so far haven't had issues may experience them just by moving the PC around a bit, where the moving GPU might slightly loosen the screw and cause it to short
You know, i'd drill out the hole larger and then fill it with epoxy and drill back out to screw size - the PCB material is epoxy-glassfibre to begin with. So the board would essentially be guaranteed to spec then. There can't be anything actually useful in that part of the board, probably just an overhang of a layer pour. But then that's me, i'm a crazy person. They really should remanufacture the boards and mail out replacements, it's not all that expensive.
But it does take time, a few months, which is something they don't have at the moment. You can get these nylon screws at moment's notice from numerous suppliers around the world, so i can't exactly blame them for taking immediate action and not sitting on their arse. Maybe they could do a second pass and now also send everyone a new riser a month or two later. AND MARK IT SOMEHOW so people don't confuse them and can determine whether they have a good riser in their PC or not.
@@defeqel6537 PC is a machine with vibration inducing parts - fans, inductors. Movement is inevitable.
Yeah. This is why some recalls specify to remove and disable the defective equipment so nobody tries to use it later.
"bad heat condition"
yeah, that's one way to describe a FIRE!
(He was reffering to the cause that lead to fire shortly after.)
Writing an email..- subject... FIRE...
Insufficient heat dissipation. Just water cool the PCI-E riser cable!
I remember years ago about the time that the DX4-100 was released when a machine came into the shop from a distressed end user that had caught fire in an unusual way, 2 wires had burned the insulation off in the centre of the floppy drive ribbon from the Vesa IO card to the back of the floppy drive, the cable seemed to be the only casualty - on inspection it was discovered that the end user missed a pin on the smaller molex power connector when refitting after carrying out an upgrade, even though it was the data cable that burned - strangely enough the IO card and the floppy drive survived -
Steve and company, great video, love the work you do bringing the techie side of computers to the masses. I do have one issue with one assumption regarding the screw placement and the motion of the thread on the screw. The thread itself would not move up and down as shown in the video. Simply look at the inside of a nut and notice it would be in a fixed position, same goes for the thread on the screw as it moves in and out of the screw-hole. A possible reason why some systems are affected and not others maybe as a unit is assembled the screw could be inserted in such a way that the pitch of the threads catches the trace on the PCB. One could assume any number of possible ways a screw can be inserted may result in any number of possible ways the thread catches the PCB trace. I do think it kind of sloppy that the PCB was designed in such a way that the result can be so dangerous. Anyhow, keep up the good work!!
"had this for 5 years, surprised that it still works" he says referring to the "fire extinguisher" as the fire is already started and he is trying to put it out lol
It works because it was not made in the UK
They have actual serious fire extinguishers too. They took the little spray bottle extinguisher because it doesn't require a lot of cleanup and is cheap to replenish. And because it was due for replacement anyway.
Yeah, I have real fire extinguishers as part of standard office requirements. I just didn't want to use one that important for a fire I had full control of. I mean, we could have just thrown the case in the parking lot if we needed to, or picked up the real fire extinguisher.
@@tj71520 ????
@@GamersNexus Yeetus yeetus, fire deletus?
NZXT could have advertise this to be the first case with real time plasma ray tracing technology.
“No GPUs are buyable, so let’s light things on fire, that works, right?”
That's not true. Right now I can see a 1GB GT 710 for sale at MSRP. Better buy it now though before those too sell out.
There's couple 3090 for 2000-2300€ around here... That counts right? Also some 3070 for 800-1200€ :/
Yay for availability of gpus. :D
Steve: What’s a great idea for our long haired staff to undertake?
Staff: Let’s set something on fire!
Sorry Steve.... when i first saw the Vid/Thumbnail I assumed it to be clickbait - since this is YT after all. Coming across the story now I realise I should have never doubted your integrity. Thank you for keeping it real.
Silly me, I'm used to best practices being have a keepout zone for traces near mounting point unless you are intentionally make a connection... :P
In fact, your PCB software will yell at you, your fab will yell at you and the test engineers will yell at you. I don't know how something like this can make it into production.
@@squelchedotter Unless, big brain moment in here... You don't add any mounting holes in the design soft since your customer was like, no need, no need - we sort it out ourselves...
@@squelchedotter as a last minute change that bypasses most of the people and machinery in the chain for example. You can easily edit the Exellon drill file by hand.
Power planes might not have the same restrictions as traces.
@@konzetsu6068 Best practices is to leave a keep out zone for all layers. For exactly this type of problem. Even KiCad will do this automatically unless you override the settings.
15:50 "you live another day"
GPU wipes off bead of sweat🤣
Patrick Stone
"Patrick in Chief"
_One more_ and it'd be a Patrick hat trick.
@@bluephreakr as they could all change their names without a problem that is simply achievable
@@bluephreakr a... Pat-trick....
So the thing is with screws, the thread doesn't move up or down at a particular connection point. It slides in the slot, but that slot remains constant. The CGI video you show is not accurate of how a screw moves into it's threaded hole. There is some movement, technically, on an extremely small scale that is probably indiscernible to the unaided human eye. i will agree that the minute movement of simply turning the screw triggers the short, as any pressure moves the screw in an incredibly small amount up or down in the hole. The original screw is painted so not only does the screw need to move (and not just rotate) in the hole but that has to coincide with stripping of paint off that screw. It also appears, to me, that a layer inside the PCB burns up, through the entirety of the PCB. The fire itself seems to be a byproduct of the short where the plastic catches fire, but I'm fairly certain that that particular layer of material inside the PCB burns (albeit with no flames, but things can burn without flame).
Its great all the staff are chill and nobody yelps or squeaks.
So, what you're saying is, the H1 case name stands for HOT ONE!!! Definitely lives up to the name.
This needs to be seened by more people. A recall is in order.
there already was a recall. this video is like a month late
"seen" is already past tense...
"It catches fire"
[Changes screws]
"I don't think that's how you stop fire..."
Technically it is, but it's a bit like opening the door to let the gas out.
I mean it is a way to stop it, it's just not a very good long term solution.
Maybe I'm just old, but from what I've seen PCBs usually have a larger-than-a-screw-head safe zone around screw holes.
Nah, someone royally fucked this up, best practice is to make sure the screw is grounded across all layers on the pcb like it's a via.
I... think this might be the episode to show to my dad, the retired electrical engineer who worked his way up to become a global manager for risk management of medical devices at a certain large safety science certification company. He'll appreciate it! :D
I can not believe that we already got the case for "The Dissapointment PC 2021" and is not even February yet lol